I am seeing the same symptom as Sim. Seems to have popped up after upgrading from Debian 4 (2.6.18-6-686-BIGMEM) to Debian 5 (2.6.26-2-686-BIGMEM). I have a bout 25 VMs running ESX 4 and 3.5 that are all running debian kernel 2.6.26-2. They are all 2 CPU VMs runnning on 2, 4, and 8 way systems. One core locks up @100% after about 10-20 days of runtime. In some cases the problem occurs within a day or so in others it takes a month. The guest itself does not show 100% utilization in top. The system does come out of this "lockup" situation after about 5 hours and the system resumes as usual.
I am currently in the process of downgrading the kernel on each of these guests back to 2.6.18 to see if I can resolve this. Unfortunately that means loosing many of the VIM paravirt features that were added since 2.6.18. I don't understand why this issue was closed as the original poster, Sim, did not indicate it was addressed. Unless Simon Boggis is the same poster as Sim which I don't think he is. -- Server 8.04 LTS: soft lockup - CPU#1 stuck for 11s! [bond1:3795] - bond - bond0 https://bugs.launchpad.net/bugs/245779 You received this bug notification because you are a member of Ubuntu Bugs, which is a direct subscriber. -- ubuntu-bugs mailing list ubuntu-bugs@lists.ubuntu.com https://lists.ubuntu.com/mailman/listinfo/ubuntu-bugs